English | 한국어 | 日本語 | 中文 | Deutsch | Français

Looking for Codex CLI? → my-codex — same orchestration in native TOML format

my-claude

All-in-one agent harness for Claude Code. One plugin, 200+ agents ready.

Boss auto-discovers every agent, skill, and MCP tool at runtime,
then routes your task to the right specialist. No config files. No boilerplate.

Installation

For Humans

git clone --depth 1 https://github.com/sehoon787/my-claude.git /tmp/my-claude
bash /tmp/my-claude/install.sh
rm -rf /tmp/my-claude

Or install as a Claude Code plugin first, then run the companion installer:

# Inside a Claude Code session:
/plugin marketplace add sehoon787/my-claude
/plugin install my-claude@my-claude

# Then install companion tools:
git clone --depth 1 https://github.com/sehoon787/my-claude.git /tmp/my-claude
bash /tmp/my-claude/install.sh
rm -rf /tmp/my-claude

For AI Agents

Read https://raw.githubusercontent.com/sehoon787/my-claude/main/AI-INSTALL.md and follow every step.

Open-Source Tools Used

1. Oh My Claude Code (OMC)

An agent harness dedicated to Claude Code. 18 specialist agents (architect, debugger, code reviewer, security reviewer, etc.) divide work by role, and magic keywords like autopilot: activate automatic parallel execution.

2. Oh My OpenAgent (omo)

A multi-platform agent harness. Bridges to the Claude Code ecosystem via claude-code-agent-loader and claude-code-plugin-loader. Automatically routes across 8 providers (Claude, GPT, Gemini, etc.) by category. The 9 agents in this repository are adaptations of omo agents in Claude Code standalone .md format.

3. Andrej Karpathy Skills

The 4 AI coding behavioral guidelines proposed by Andrej Karpathy (Think Before Coding, Simplicity First, Surgical Changes, Goal-Driven Execution). Included in CLAUDE.md and always active across all sessions.

4. Everything Claude Code (ECC)

A comprehensive framework providing 67 skills + 17 agents + 45 commands + language-specific rules. Automates repetitive development patterns with slash commands like /tdd, /plan, /code-review, and /build-fix.

5. Anthropic Official Skills

The official agent skills repository provided directly by Anthropic. Enables specialist tasks such as PDF parsing, Word/Excel/PowerPoint document manipulation, and MCP server creation.

6. Agency Agents

A library of 164 business specialist agent personas. Provides specialist perspectives in business contexts beyond technical roles — UX architects, data engineers, security auditors, QA managers, and more.

7. gstack

A sprint-process harness by Garry Tan with 27 skills. Provides browser-based QA testing (/qa), code review with scope-drift detection (/review), security auditing (/cso), and a full Plan→Review→QA→Ship deployment workflow. Includes a compiled Playwright browser daemon for real-browser testing.

8. VibeProxy

OAuth-based API proxy that uses existing AI subscriptions (ChatGPT Plus, Gemini, Claude) to access GPT/Gemini/Claude models without separate API keys. Supports both Anthropic and OpenAI API formats. Integrated via optional vibeproxy-setup.sh during installation. Model routing maps Claude model tiers (opus/sonnet/haiku) to equivalent vendor models.

How Boss Works

Boss is the meta-orchestrator at the core of my-claude. It never writes code — it discovers, classifies, matches, delegates, and verifies.

User Request
     │
     ▼
┌─────────────────────────────────────────────┐
│  Phase 0 · DISCOVERY                        │
│  Scan agents, skills, MCP, hooks at runtime │
│  → Build live capability registry           │
└──────────────────────┬──────────────────────┘
                       ▼
┌─────────────────────────────────────────────┐
│  Phase 1 · INTENT GATE                      │
│  Classify: trivial | build | refactor |     │
│  mid-sized | architecture | research | ...  │
│  → Counter-propose skill if better fit      │
└──────────────────────┬──────────────────────┘
                       ▼
┌─────────────────────────────────────────────┐
│  Phase 2 · CAPABILITY MATCHING              │
│  P0: gstack skill (if installed)            │
│  P1: Exact skill match                      │
│  P2: Specialist agent (200+)               │
│  P3: Multi-agent orchestration              │
│  P4: General-purpose fallback               │
└──────────────────────┬──────────────────────┘
                       ▼
┌─────────────────────────────────────────────┐
│  Phase 3 · DELEGATION                       │
│  6-section structured prompt to specialist  │
│  TASK / OUTCOME / TOOLS / DO / DON'T / CTX  │
└──────────────────────┬──────────────────────┘
                       ▼
┌─────────────────────────────────────────────┐
│  Phase 4 · VERIFICATION                     │
│  Read changed files independently           │
│  Run tests, lint, build                     │
│  Cross-reference with original intent       │
│  → Retry up to 3× on failure               │
└─────────────────────────────────────────────┘

Priority Routing

Boss cascades every request through a priority chain until the best match is found:

Model Routing

3-Phase Sprint Workflow

For end-to-end feature implementation, Boss orchestrates a structured sprint:

Phase 1: DESIGN         Phase 2: EXECUTE        Phase 3: REVIEW
(interactive)            (autonomous)             (interactive)
─────────────────────   ─────────────────────   ─────────────────────
User decides scope      ralph runs execution    Compare vs design doc
Engineering review      Auto code review        Present comparison table
Confirm "design done"   Architect verification  User: approve / improve

Architecture

┌─────────────────────────────────────────────────────┐
│                    User Request                       │
└───────────────────────┬─────────────────────────────┘
                        ▼
┌─────────────────────────────────────────────────────┐
│  Boss · Meta-Orchestrator (Opus)                      │
│  Discovery → Classification → Matching → Delegation  │
└──┬──────────┬──────────┬──────────┬─────────────────┘
   │          │          │          │
   ▼          ▼          ▼          ▼
┌──────┐ ┌────────┐ ┌────────┐ ┌────────┐
│ P3a  │ │  P3b   │ │  P3c   │ │  P1/P2 │
│Direct│ │Sub-orch│ │ Agent  │ │ Skill/ │
│2-4   │ │Sisyphus│ │ Teams  │ │ Agent  │
│agents│ │Atlas   │ │  P2P   │ │ Direct │
└──────┘ │Hephaes│ └────────┘ └────────┘
         └────────┘
┌─────────────────────────────────────────────────────┐
│  Behavioral Layer                                     │
│  Karpathy Guidelines · ECC Rules (87) · Hooks (7)    │
├─────────────────────────────────────────────────────┤
│  Specialist Agents (200+)                             │
│  OMO 9 · OMC 19 · Agency Eng. 26 · Superpowers 1    │
│  + 136 domain packs (on-demand)                       │
├─────────────────────────────────────────────────────┤
│  Skills (200+)                                        │
│  ECC 180+ · OMC 36 · gstack 40 · Superpowers 14     │
│  + Core 3 · Anthropic 14+                             │
├─────────────────────────────────────────────────────┤
│  MCP Layer                                            │
│  Context7 · Exa · grep.app                            │
└─────────────────────────────────────────────────────┘

What's Inside

ln -s ~/.claude/agent-packs/marketing/*.md ~/.claude/agents/

MCP Servers (3) + Hooks (7)

MCP Servers

Server	Purpose	Cost
Context7	Real-time library documentation	Free
Exa	Semantic web search	Free 1k req/month
grep.app	GitHub code search	Free

Behavioral Hooks

Hook	Event	Behavior
Session Setup	SessionStart	Auto-detects missing tools + injects Briefing Vault context
Delegation Guard	PreToolUse	Blocks Boss from directly modifying files
Agent Telemetry	PostToolUse	Logs agent usage to agent-usage.jsonl
Subagent Verifier	SubagentStop	Forces independent verification + logs to Briefing Vault
Completion Check	Stop	Runs profile fallback + guards /boss-briefing execution
Teammate Idle Guide	TeammateIdle	Prompts leader on idle teammates
Task Quality Gate	TaskCompleted	Verifies deliverable quality
Vault Reminder	UserPromptSubmit	Suggests /boss-briefing after 5+ messages

Briefing Vault

Obsidian-compatible persistent memory. Every project maintains a .briefing/ directory that auto-populates across sessions.

.briefing/
├── INDEX.md                          ← Project context (auto-created once)
├── state.json                        ← Session metadata, counters, lastVaultSync (auto-managed)
├── sessions/
│   ├── YYYY-MM-DD-<topic>.md        ← AI-written session summary (enforced)
│   └── YYYY-MM-DD-auto.md           ← Auto-generated scaffold (git diff, agent stats)
├── decisions/
│   └── YYYY-MM-DD-<decision>.md     ← AI-written decision record (enforced)
├── learnings/
│   ├── YYYY-MM-DD-<pattern>.md      ← AI-written learning note
│   └── YYYY-MM-DD-auto-session.md   ← Auto-generated scaffold (agents, files)
├── references/
│   └── auto-links.md                ← Auto-collected URLs from web searches
├── archives/                         ← PARA: completed/inactive notes (flat)
├── wiki/                             ← LLM-wiki: concept pages
│   └── _schema.md
├── agents/
│   ├── agent-log.jsonl              ← Subagent execution telemetry
│   └── YYYY-MM-DD-summary.md        ← Daily agent usage breakdown
└── persona/
    ├── profile.md                   ← Agent affinity stats (auto-updated)
    ├── suggestions.jsonl            ← Routing suggestions (auto-generated)
    └── rules/                       ← Workflow pattern rules (workflow-*.md)

Sub-Vaults

Knowledge Management (v2)

BriefingVault v2 integrates three knowledge management methodologies:

Claude Code session-end hooks automatically:

• Suggest archiving notes older than 30 days
• Propose wiki pages for frequently mentioned concepts
• Generate unique Zettelkasten IDs for new notes

Session-Specific Diffs

At session start, the current git HEAD is saved to state.json (sessionStartHead field). For non-git projects, a YYYY-MM-DD:cwd identifier is used instead. At session end, diffs are calculated relative to this saved point — showing only changes from the current session, not accumulated uncommitted changes from previous sessions.

Using with Obsidian

1. Open Obsidian → Open folder as vault → select .briefing/
2. Notes appear in graph view, linked by [[wiki-links]]
3. YAML frontmatter (date, type, tags) enables structured search
4. Timeline of decisions and learnings builds automatically over sessions

/boss-briefing

Run /boss-briefing during or at the end of a session to:

• Sync vault: Update profile.md, INDEX.md, and agent summaries
• Detect workflow patterns: Analyze temporal agent call sequences across sessions
• Recover from gaps: Generate recovery summaries if days have passed since the last session
• Propose persona rules: Suggest workflow-based routing preferences (not just frequency)
• Validate session notes: Check that today's session has a proper summary

The Stop hook checks whether /boss-briefing has run today. If not, it blocks session end with a reminder. The existing stop-profile-update.js continues to run as a fallback.

Upstream Open-Source Sources

my-claude bundles content from 5 MIT-licensed upstream repositories via git submodules:

#	Source	What It Provides
1	oh-my-claudecode — Yeachan Heo	19 specialist agents + 36 skills. Claude Code multi-agent harness with autopilot, ralph, team orchestration.
2	oh-my-openagent — code-yeongyu	9 OMO agents (Sisyphus, Atlas, Oracle, etc.). Multi-platform agent harness bridging Claude, GPT, Gemini.
3	everything-claude-code — affaan-m	180+ skills + 87 rules across 14 languages. Comprehensive dev framework with TDD, security, and coding standards.
4	agency-agents — msitarzewski	26 engineering agents (always loaded) + 136 domain agent-packs across 12 categories.
5	gstack — garrytan	40 skills for code review, QA, security audit, deployment. Includes Playwright browser daemon.
6	superpowers — Jesse Vincent	14 skills + 1 agent covering brainstorming, TDD, parallel agents, and code review.
7	anthropic/skills — Anthropic	14+ official skills for PDF, DOCX, PPTX, XLSX, and MCP builder.
8	andrej-karpathy-skills — forrestchang	4 AI coding behavioral guidelines (Think Before Coding, Simplicity First, Surgical Changes, Goal-Driven Execution).

GitHub Actions

my-claude Originals

Features built specifically for this project, beyond what upstream sources provide:

Bundled Upstream Versions

Linked via git submodules. Pinned commits tracked natively by .gitmodules.

Contributing

Issues and PRs are welcome. When adding a new agent, add a .md file to agents/core/ or agents/omo/ and update SETUP.md.

Credits

Built on the work of: oh-my-claudecode (Yeachan Heo), oh-my-openagent (code-yeongyu), everything-claude-code (affaan-m), agency-agents (msitarzewski), gstack (garrytan), superpowers (Jesse Vincent), anthropic/skills (Anthropic), andrej-karpathy-skills (forrestchang).

License

MIT License. See the LICENSE file for details.